To'liq siljishli model nimani anglatadi?

To'liq siljishli model nimani anglatadi?

To'liq konvolyutsiyali modeli faqat konstruktiv qatlamlarga ega bo'lgan (Batch-Norm va Activation bilan) model bo'lganmi yoki yo'qmi: max-pool, fully-connected and other layers?

If it is true, then why Yolo v2 neural network for object detection is named Fully-Convolutional model if it uses Max-pool layers: https://arxiv.org/pdf/1612.08242v1.pdf

Jadval 2: YOLO dan YOLOv2gacha bo'lgan yo'l. Ko'rsatilgan dizayni ko'pchiligi   qarorlar MAPning sezilarli darajada oshishiga olib keldi. Ikki istisnosiz    zanjirli qutilarga ega va yangi tarmoqdan foydalangan holda to'liq konvolyutsiyali tarmoqqa o'tish .

enter image description here


Also why Fully Convolutional Networks for Semantic Segmentation is named Fully Convolutional if it has Max-pool layers: https://people.eecs.berkeley.edu/~jonlong/long_shelhamer_fcn.pdf

enter image description here

3

1 javoblar

Umuman olganda, butunlay bog'langan qatlamlari bo'lmagan CNN tarmog'i To'liq Konvensional Tarmoq (FCN) deb ataladi. Bu har qanday tipdagi birlashma qatlamini, ommaviy me'yorini, qoldirib ketishni va boshqalarni o'z ichiga olishi mumkin.

3
qo'shib qo'ydi